Thesis (M.S.)--University of Wisconsin--Madison. / Typescript. eContent provider-neutral record in process. Description based on print version record. Includes bibliographical references (leaves 98-104).

Thesis (M.Soc.Sc.)--University of Hong Kong, 1991. / Also available in print.

This special volume presents a select number of psychometric techniques, many of them original, and their implementation in R packages. (authors' abstract)

The purpose of this study was to investigate the construct validity of key concepts from three prominent instruments in the family therapy field (FACES III, SFI, and FES). Through analysis of the shared variance between the items and scales which make up these instruments the convergent and divergent validity of these measures was established. Specific objectives included the assessment of (1) psychometric properties of these instruments, (2) the relative importance of the different theoretical constructs (operationalized in the scales), (3) a higher order variable working throughout each construct. / Three hundred thirty-two college students participated in this study. Each subject filled out a packet containing a demographic questionnaire and each of the three instruments. The findings were analyzed through a variety of methods. Reliabilities and factor analyses were calculated for each scale within each instrument. The instruments were then combined and additional factor analyses were run on both the item and scale level. Pearson Product Correlations were calculated between each scale as well. Finally a LISREL analysis was attempted for each instrument. / The findings of this study revealed a sense of belonging as the primary dimension from which these subjects viewed their family. The dimensions of Organization, Intellectual-Cultural Orientation, and Moral-Religious Emphasis fell out as well. While several scales of the SFI and FES loaded on this central dimension, many of the dimensions researchers believed to be important to family functioning were not discerned by these family members. Thus a single higher order construct operating throughout all the scales was not found. Questions regarding the utility of these instruments were raised and implications for future research were made. / Source: Dissertation Abstracts International, Volume: 49-12, Section: B, page: 5565. / Major Professor: Mary Hicks. / Thesis (Ph.D.)--The Florida State University, 1988.

The purpose of this study was to determine whether two parallel tests could be equated effectively (i.e., minimal equating error) using the multidimensional two-parameter IRT model (M2PL). Equatings were performed on both simulated and real data sets using the common-examinee design and two abilities for each examinee (i.e., two dimensions). Item and ability parameter estimates were found using the Multidimensional Item Response Theory Estimation (MIRTE) program developed by Carlson at ACT. The amount of equating error was evaluated by a comparison of the contour plots, response surfaces, and the mean difference and absolute mean difference between the true scores and ability estimates found on both tests for the common examinees used in the equating. / The results indicated that effective equating was possible with the techniques used in this study when true scores, the response surfaces, and the contour plots were compared. When the equating of the ability estimates was examined, unsatisfactory results were found. It was concluded that tests containing more than 40 items would be needed for stable ability estimates. / Source: Dissertation Abstracts International, Volume: 49-06, Section: B, page: 2409. / Major Professor: John R. Hills. / Thesis (Ph.D.)--The Florida State University, 1988.

Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 2008. / Source: Dissertation Abstracts International, Volume: 69-11, Section: B, page: 7185. Adviser: Hua-Hua Chang. Includes bibliographical references (leaves 84-91) Available on microfilm from Pro Quest Information and Learning.

Thesis--University of Oslo. / Bibliography: p. [58]

Thesis (Ph. D.)--University of Missouri-Columbia, 2003. / Typescript. Vita. Includes bibliographical references (leaves 88-95). Also available on the Internet.

Thesis (Ph. D.)--University of Missouri-Columbia, 2003. / Typescript. Vita. Includes bibliographical references (leaves 88-95). Also available on the Internet.

The majority of data used in outcome research has until recently been based primarily on treatment provider observations of the patient. Estimating the effect of rehabilitation treatment on functional ability is increasingly relying on patient self-report ratings. Controlling for measurement bias associated with the subjective nature of this type of data was handled by putting self-report and clinician responses to items on an 11-item functional ability instrument into the same metric using a Rasch model, allowing for an accurate picture of whether and how their perceptions differ. For this study, the traditionally clinician-rated instrument was modified to accommodate self-report ratings. The questionnaire was verbally administered during individualized discharge interviews to 112 visually-impaired geriatric veterans whose responses were then paired with clinician ratings. Clinician ratings replicate previous findings which provided a stable foundation against which the self-report ratings were compared. Fit statistics for self-report ratings demonstrate unidimensionality of the assumed functional ability construct. Construct validity, or to what extent clinicians and patients agree on what it is that is being measured (i.e., functional ability), was supported by a moderate correlation between person ability estimates. More informative, however, was the agreement on item difficulty estimates. While six of the 11 items had roughly the same difficulty estimates across the two rating samples, 5 others lacked agreement. Two possible reasons for this incongruity are (1) patient expectations about what they should be able to do by the end of rehabilitation are not realized and (2) exemplars used in eliciting self-report ratings are not as specific as those used by clinicians. The single-factor structure was also analyzed by a confirmatory factor analysis. Although the model did not meet conventional criteria of .95 for the CFI, it was the most parsimonious solution for both sets of ratings as demonstrated by a multisample analysis. These analyses indicated that self-perceived functional ability ratings may be used in place of clinician ratings on the majority of the FAST items. Until the remaining items are modified, at this point, self-report ratings may, at best, supplement, but not be used in place of, clinician ratings.